Metal Fabrication Quoting: Promise Dates From a Real Capacity Simulation

Metal fabrication quoting software that runs a real capacity simulation lets a fab shop quote lead time and cost from an actual schedule of the quoted job against current load, and EDGEBIC by User Solutions does this in memory without touching the live plan. For a shop juggling press brakes, lasers, welders, and paint, the quote is a promise about a queue you cannot see by eye. This post is about replacing the gut-feel date with a simulated one, and reading the cost that comes with it.

The general idea of loading a job against finite capacity is covered in what is production scheduling, and the choice between quoting forward from today or backward from a due date is forward versus backward scheduling. Here we apply both to the estimator's desk. One honesty note up front: this is quote simulation and what-if promise dates, a schedule you run and read. It is not a live commit-to-promise transaction against reserved capacity.

Why the gut-feel quote costs you both ways

A fabrication quote is really two numbers under pressure: how much, and how soon. Get the soon wrong on the high side and you lose the job to a shop that promised faster. Get it wrong on the low side and you blow the date, eat expedite costs, and burn the customer. The honest answer depends entirely on how backed up your bottleneck is the week the job would run, and that changes daily as new orders land.

An estimating spreadsheet cannot see that queue. It carries fixed hours and a fixed lead time, so every quote assumes an empty shop. The real shop has a four-day wait on the brake and a welder that is booked solid, and none of that is in the number you send. Fab shops feel this acutely because routings are multi-step and the constraint moves: this week it is the laser, next week it is paint. Understanding which resource is binding is the subject of bottleneck identification, and a good quote has to respect it.

How the quote simulation works

When you build a quote, the system creates a temporary order for the quoted product and quantity, then runs the full finite-capacity scheduler against your current shop load. Two things make this safe to do all day. It runs in memory, and it writes nothing to the live schedule. The temporary order exists only long enough to be scheduled and is removed afterward, so your production plan is never disturbed and you can run twenty simulations back to back.

The simulation respects the same shift calendars, machine capacity, routing sequence, and setup times as a real scheduling run. It differs in exactly one way: the result is read and discarded rather than saved. From the simulated schedule the system reads back the estimated start and end dates, which gives you a lead time that reflects the queue the job would actually join.

Reading the cost off the schedule

Cost comes from the same simulated schedule. The system sums the work hours on each work center in the routing and multiplies by that center's labor rate. It adds material cost, either from material steps in the routing or, as a fallback, from the product's unit cost times the quantity. Then it applies your markup to derive a unit price.

Take a 200-piece order with two steps: a milling step at 8 hours per unit and an assembly step at 2 hours per unit, with mill at 55 dollars an hour, assembly at 30, and 12 dollars of material per piece.

That rolls up to 102,400 dollars of cost. At a 25 percent markup the unit price is 640 dollars, for a total of 128,000 dollars, and the simulation also gives you the estimated finish date to compare against the customer's request. Critically, the hours figure is real work content, not wall-clock span, so a job that crosses a weekend is not charged for the idle days.

When you know the number needs an adjustment the routing cannot see, tooling wear or an unusual scrap rate, a planner can override the rolled-up cost with a fixed figure. The system still stores the detailed labor and material breakdown, so later you can compare the estimate to actuals line by line.

Testing a customer's requested date directly

The most useful move for a fab shop is the backward quote. Instead of asking "when would this finish if we start now," a backward quote asks "can we finish by the date the customer wants." The simulation right-aligns the job to end at or before the requested date, starting as late as feasible.

If the work fits, the estimated end lands on or before the requested date and you have your answer. If it does not fit, the simulation falls back to a forward pass and returns a later completion, which is the system telling you the date is not achievable and showing when the job would actually finish. One run answers a question that a spreadsheet cannot answer at all. The direction concept, and why the earliest-start floor stays at "now" rather than the due date, is covered in forward versus backward scheduling.

From accepted quote to scheduled order

When the customer accepts, you convert the quote to a manufacturing order in one step. The conversion carries the cost fields, the markup, the unit price, the customer link, and the scheduling direction to the new order. A backward quote becomes a backward order and stays that way until actuals or a fixed constraint force it forward. Nothing is re-keyed, and the stored estimate becomes the baseline for quote-versus-actual variance once the job runs and actuals flow in.

That closed loop is the payoff. The estimate, the schedule, and the eventual variance all come from one model of your shop, not from a spreadsheet that drifts away from the plan the day after you save it.

Quoting alongside the rest of the fab shop plan

A good quote respects the same shop realities the production plan does. If the quoted routing runs steps in parallel to compress lead time, the simulation sees it, the way parallel work centers for fabrication do on the live plan. If it uses lot streaming to overlap operations, that overlap is in the simulated dates too, as in lot streaming for fabrication. And the whole estimate is only as trustworthy as the shop's bottleneck picture, which is exactly what bottleneck scheduling for fabrication keeps honest.

This is the same estimating discipline User Solutions has brought to manufacturers since 1991: promise dates from a real capacity model rather than a hopeful number, the kind of grounding that took GE Railcar from 30 percent on-time to 90 percent.

Put a real RFQ in front of it. Bring a routing and a requested date to a demo and watch the simulation return a lead time and a price your shop can stand behind. When you are ready to feed it your product and order data, EDGEBIC reads it through flexible import and export masks.